@houtini/seo-audit-console 0.2.0

package/dist/core/dataStorage.d.ts

@@ -0,0 +1,42 @@
+/** How many recent audit runs / crawl snapshots clear-crawl-history keeps. */
+export declare const KEEP_RECENT = 5;
+export interface PropertyStorage {
+    file: string;
+    siteUrl: string | null;
+    bytes: number;
+    searchAnalytics: number;
+    pages: number;
+    links: number;
+    pageSnapshots: number;
+    findings: number;
+    lastSynced: string | null;
+    lastCrawl: string | null;
+}
+export interface StorageSummary {
+    dataDir: string;
+    properties: PropertyStorage[];
+    caches: {
+        file: string;
+        bytes: number;
+    }[];
+    other: {
+        file: string;
+        bytes: number;
+    }[];
+    reportsBytes: number;
+    totalBytes: number;
+}
+export declare function storageSummary(dataDir: string, siteUrl?: string): StorageSummary;
+export declare const fmtBytes: (n: number) => string;
+export interface PruneResult {
+    ok: boolean;
+    refused?: string;
+    action: string;
+    siteUrl: string;
+    detail: Record<string, unknown>;
+}
+export declare function pruneProperty(dataDir: string, siteUrl: string, action: 'vacuum' | 'clear-crawl-history' | 'delete-property', confirm: boolean, runningJobs: {
+    id: string;
+    type: string;
+}[]): PruneResult;
+//# sourceMappingURL=dataStorage.d.ts.map

package/dist/core/dataStorage.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"dataStorage.d.ts","sourceRoot":"","sources":["../../src/core/dataStorage.ts"],"names":[],"mappings":"AAWA,8EAA8E;AAC9E,eAAO,MAAM,WAAW,IAAI,CAAC;AAI7B,MAAM,WAAW,eAAe;IAC9B,IAAI,EAAE,MAAM,CAAC;IACb,OAAO,EAAE,MAAM,GAAG,IAAI,CAAC;IACvB,KAAK,EAAE,MAAM,CAAC;IACd,eAAe,EAAE,MAAM,CAAC;IACxB,KAAK,EAAE,MAAM,CAAC;IACd,KAAK,EAAE,MAAM,CAAC;IACd,aAAa,EAAE,MAAM,CAAC;IACtB,QAAQ,EAAE,MAAM,CAAC;IACjB,UAAU,EAAE,MAAM,GAAG,IAAI,CAAC;IAC1B,SAAS,EAAE,MAAM,GAAG,IAAI,CAAC;CAC1B;AAED,MAAM,WAAW,cAAc;IAC7B,OAAO,EAAE,MAAM,CAAC;IAChB,UAAU,EAAE,eAAe,EAAE,CAAC;IAC9B,MAAM,EAAE;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;IAC1C,KAAK,EAAE;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;IACzC,YAAY,EAAE,MAAM,CAAC;IACrB,UAAU,EAAE,MAAM,CAAC;CACpB;AA0BD,wBAAgB,cAAc,CAAC,OAAO,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,MAAM,GAAG,cAAc,CAiDhF;AAED,eAAO,MAAM,QAAQ,GAAI,GAAG,MAAM,KAAG,MAKpC,CAAC;AAEF,MAAM,WAAW,WAAW;IAC1B,EAAE,EAAE,OAAO,CAAC;IACZ,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,MAAM,EAAE,MAAM,CAAC;IACf,OAAO,EAAE,MAAM,CAAC;IAChB,MAAM,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CACjC;AAED,wBAAgB,aAAa,CAC3B,OAAO,EAAE,MAAM,EACf,OAAO,EAAE,MAAM,EACf,MAAM,EAAE,QAAQ,GAAG,qBAAqB,GAAG,iBAAiB,EAC5D,OAAO,EAAE,OAAO,EAChB,WAAW,EAAE;IAAE,EAAE,EAAE,MAAM,CAAC;IAAC,IAAI,EAAE,MAAM,CAAA;CAAE,EAAE,GAC1C,WAAW,CAsEb"}

package/dist/core/dataStorage.js

@@ -0,0 +1,193 @@
+/**
+ * dataStorage — data-hygiene: what's on disk per property (sizes, row counts, last
+ * sync/crawl), plus explicit prune actions (vacuum / clear-crawl-history /
+ * delete-property). Destructive actions require confirm:true and refuse loudly,
+ * naming exactly what would be deleted. Never touches a DB while a job is running.
+ */
+import Database from 'better-sqlite3';
+import { readdirSync, statSync, existsSync, rmSync } from 'node:fs';
+import path from 'node:path';
+import { dbPathFor, sanitizeProperty } from './paths.js';
+/** How many recent audit runs / crawl snapshots clear-crawl-history keeps. */
+export const KEEP_RECENT = 5;
+const CACHE_DB_NAMES = new Set(['dataforseo-cache.db', 'wikidata-cache.db']);
+const sizeWithSidecars = (dbFile) => {
+    let total = 0;
+    for (const suffix of ['', '-wal', '-shm']) {
+        try {
+            total += statSync(dbFile + suffix).size;
+        }
+        catch { /* absent */ }
+    }
+    return total;
+};
+const dirSize = (dir) => {
+    let total = 0;
+    try {
+        for (const f of readdirSync(dir, { withFileTypes: true })) {
+            const p = path.join(dir, f.name);
+            if (f.isDirectory())
+                total += dirSize(p);
+            else {
+                try {
+                    total += statSync(p).size;
+                }
+                catch { /* raced */ }
+            }
+        }
+    }
+    catch { /* absent */ }
+    return total;
+};
+const safeCount = (db, table) => {
+    try {
+        return db.prepare(`SELECT COUNT(*) n FROM "${table}"`).get().n;
+    }
+    catch {
+        return 0;
+    }
+};
+export function storageSummary(dataDir, siteUrl) {
+    const properties = [];
+    const caches = [];
+    const other = [];
+    let files = [];
+    try {
+        files = readdirSync(dataDir).filter(f => f.endsWith('.db'));
+    }
+    catch { /* dataDir absent */ }
+    const wantStem = siteUrl ? sanitizeProperty(siteUrl) : null;
+    for (const file of files.sort()) {
+        const full = path.join(dataDir, file);
+        const bytes = sizeWithSidecars(full);
+        if (CACHE_DB_NAMES.has(file)) {
+            caches.push({ file, bytes });
+            continue;
+        }
+        let db = null;
+        try {
+            db = new Database(full, { readonly: true, fileMustExist: true });
+            const isProperty = !!db.prepare(`SELECT 1 FROM sqlite_master WHERE type='table' AND name='property_meta'`).get();
+            if (!isProperty) {
+                other.push({ file, bytes });
+                continue;
+            }
+            const meta = db.prepare('SELECT site_url, last_synced_at FROM property_meta ORDER BY id LIMIT 1').get();
+            if (wantStem && file !== `${wantStem}.db`)
+                continue;
+            let lastCrawl = null;
+            try {
+                lastCrawl = db.prepare(`SELECT MAX(COALESCE(finished_at, started_at)) t FROM crawl_metadata`).get().t;
+            }
+            catch { /* no crawl table */ }
+            properties.push({
+                file,
+                siteUrl: meta?.site_url ?? null,
+                bytes,
+                searchAnalytics: safeCount(db, 'search_analytics'),
+                pages: safeCount(db, 'pages'),
+                links: safeCount(db, 'links'),
+                pageSnapshots: safeCount(db, 'page_snapshots'),
+                findings: safeCount(db, 'findings'),
+                lastSynced: meta?.last_synced_at ?? null,
+                lastCrawl,
+            });
+        }
+        catch {
+            other.push({ file, bytes });
+        }
+        finally {
+            db?.close();
+        }
+    }
+    const reportsBytes = dirSize(path.join(dataDir, 'reports'));
+    const totalBytes = properties.reduce((s, p) => s + p.bytes, 0)
+        + caches.reduce((s, c) => s + c.bytes, 0)
+        + other.reduce((s, o) => s + o.bytes, 0)
+        + reportsBytes;
+    return { dataDir, properties, caches, other, reportsBytes, totalBytes };
+}
+export const fmtBytes = (n) => {
+    if (n >= 1024 ** 3)
+        return `${(n / 1024 ** 3).toFixed(2)} GB`;
+    if (n >= 1024 ** 2)
+        return `${(n / 1024 ** 2).toFixed(1)} MB`;
+    if (n >= 1024)
+        return `${(n / 1024).toFixed(0)} KB`;
+    return `${n} B`;
+};
+export function pruneProperty(dataDir, siteUrl, action, confirm, runningJobs) {
+    const base = { action, siteUrl };
+    if (runningJobs.length) {
+        return {
+            ...base, ok: false, detail: { runningJobs },
+            refused: `Refusing to ${action} while ${runningJobs.length} job(s) are running (${runningJobs.map(j => `${j.type}#${j.id}`).join(', ')}) — a prune mid-sync/crawl can corrupt or race the write. Wait for them to finish and retry.`,
+        };
+    }
+    const dbFile = dbPathFor(dataDir, siteUrl);
+    if (!existsSync(dbFile)) {
+        return { ...base, ok: false, refused: `No database for ${siteUrl} at ${dbFile} — nothing to ${action}. Check list via data_storage (no args).`, detail: { dbFile } };
+    }
+    const bytesBefore = sizeWithSidecars(dbFile);
+    if (action === 'delete-property') {
+        const summary = storageSummary(dataDir, siteUrl).properties[0];
+        if (!confirm) {
+            return {
+                ...base, ok: false, detail: { dbFile, bytes: bytesBefore, rows: summary },
+                refused: `REFUSED — delete-property is destructive and needs confirm:true. This would permanently delete ${dbFile} (+ -wal/-shm), ${fmtBytes(bytesBefore)}: ` +
+                    `${summary?.searchAnalytics ?? '?'} search_analytics rows, ${summary?.pages ?? '?'} pages, ${summary?.links ?? '?'} links, ${summary?.pageSnapshots ?? '?'} snapshots, ${summary?.findings ?? '?'} findings. ` +
+                    `GSC history is re-syncable but slow to rebuild. Re-run with confirm:true to proceed.`,
+            };
+        }
+        for (const suffix of ['', '-wal', '-shm'])
+            rmSync(dbFile + suffix, { force: true });
+        return { ...base, ok: true, detail: { dbFile, bytesFreed: bytesBefore } };
+    }
+    if (action === 'vacuum') {
+        const db = new Database(dbFile);
+        try {
+            db.pragma('wal_checkpoint(TRUNCATE)');
+            db.exec('VACUUM');
+        }
+        finally {
+            db.close();
+        }
+        const bytesAfter = sizeWithSidecars(dbFile);
+        return { ...base, ok: true, detail: { bytesBefore, bytesAfter, bytesReclaimed: Math.max(bytesBefore - bytesAfter, 0) } };
+    }
+    // clear-crawl-history: drop page_snapshots + audit_runs/findings beyond the KEEP_RECENT
+    // most recent, then VACUUM. Preview first so the refusal can name what goes.
+    const db = new Database(dbFile);
+    try {
+        const keepRuns = db.prepare(`SELECT run_id FROM audit_runs ORDER BY started_at DESC LIMIT ${KEEP_RECENT}`).all().map(r => r.run_id);
+        const keepCrawls = db.prepare(`SELECT crawl_id FROM (SELECT crawl_id, MAX(captured_at) t FROM page_snapshots GROUP BY crawl_id) ORDER BY t DESC LIMIT ${KEEP_RECENT}`).all().map(r => r.crawl_id);
+        const ph = (n) => Array.from({ length: n }, () => '?').join(',') || `''`;
+        const oldRuns = db.prepare(`SELECT COUNT(*) n FROM audit_runs WHERE run_id NOT IN (${ph(keepRuns.length)})`).get(...keepRuns).n;
+        const oldFindings = db.prepare(`SELECT COUNT(*) n FROM findings WHERE run_id NOT IN (${ph(keepRuns.length)})`).get(...keepRuns).n;
+        const oldSnapshots = db.prepare(`SELECT COUNT(*) n FROM page_snapshots WHERE crawl_id NOT IN (${ph(keepCrawls.length)})`).get(...keepCrawls).n;
+        if (!confirm) {
+            db.close();
+            return {
+                ...base, ok: false, detail: { oldRuns, oldFindings, oldSnapshots, keepRecent: KEEP_RECENT },
+                refused: `REFUSED — clear-crawl-history is destructive and needs confirm:true. This would permanently delete ${oldRuns} audit runs, ${oldFindings} findings and ${oldSnapshots} page snapshots older than the ${KEEP_RECENT} most recent (change-detection history beyond that window is lost). Re-run with confirm:true to proceed.`,
+            };
+        }
+        db.transaction(() => {
+            db.prepare(`DELETE FROM findings WHERE run_id NOT IN (${ph(keepRuns.length)})`).run(...keepRuns);
+            db.prepare(`DELETE FROM audit_runs WHERE run_id NOT IN (${ph(keepRuns.length)})`).run(...keepRuns);
+            db.prepare(`DELETE FROM page_snapshots WHERE crawl_id NOT IN (${ph(keepCrawls.length)})`).run(...keepCrawls);
+        })();
+        db.pragma('wal_checkpoint(TRUNCATE)');
+        db.exec('VACUUM');
+    }
+    finally {
+        if (db.open)
+            db.close();
+    }
+    const bytesAfter = sizeWithSidecars(dbFile);
+    return {
+        ...base, ok: true,
+        detail: { bytesBefore, bytesAfter, bytesReclaimed: Math.max(bytesBefore - bytesAfter, 0), keepRecent: KEEP_RECENT },
+    };
+}
+//# sourceMappingURL=dataStorage.js.map

package/dist/core/dataStorage.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"dataStorage.js","sourceRoot":"","sources":["../../src/core/dataStorage.ts"],"names":[],"mappings":"AAAA;;;;;GAKG;AACH,OAAO,QAAQ,MAAM,gBAAgB,CAAC;AACtC,OAAO,EAAE,WAAW,EAAE,QAAQ,EAAE,UAAU,EAAE,MAAM,EAAE,MAAM,SAAS,CAAC;AACpE,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,SAAS,EAAE,gBAAgB,EAAE,MAAM,YAAY,CAAC;AAEzD,8EAA8E;AAC9E,MAAM,CAAC,MAAM,WAAW,GAAG,CAAC,CAAC;AAE7B,MAAM,cAAc,GAAG,IAAI,GAAG,CAAC,CAAC,qBAAqB,EAAE,mBAAmB,CAAC,CAAC,CAAC;AAwB7E,MAAM,gBAAgB,GAAG,CAAC,MAAc,EAAU,EAAE;IAClD,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,KAAK,MAAM,MAAM,IAAI,CAAC,EAAE,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,CAAC;QAC1C,IAAI,CAAC;YAAC,KAAK,IAAI,QAAQ,CAAC,MAAM,GAAG,MAAM,CAAC,CAAC,IAAI,CAAC;QAAC,CAAC;QAAC,MAAM,CAAC,CAAC,YAAY,CAAC,CAAC;IACzE,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC,CAAC;AAEF,MAAM,OAAO,GAAG,CAAC,GAAW,EAAU,EAAE;IACtC,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,IAAI,CAAC;QACH,KAAK,MAAM,CAAC,IAAI,WAAW,CAAC,GAAG,EAAE,EAAE,aAAa,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC;YAC1D,MAAM,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,GAAG,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC;YACjC,IAAI,CAAC,CAAC,WAAW,EAAE;gBAAE,KAAK,IAAI,OAAO,CAAC,CAAC,CAAC,CAAC;iBACpC,CAAC;gBAAC,IAAI,CAAC;oBAAC,KAAK,IAAI,QAAQ,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;gBAAC,CAAC;gBAAC,MAAM,CAAC,CAAC,WAAW,CAAC,CAAC;YAAC,CAAC;QACnE,CAAC;IACH,CAAC;IAAC,MAAM,CAAC,CAAC,YAAY,CAAC,CAAC;IACxB,OAAO,KAAK,CAAC;AACf,CAAC,CAAC;AAEF,MAAM,SAAS,GAAG,CAAC,EAAqB,EAAE,KAAa,EAAU,EAAE;IACjE,IAAI,CAAC;QAAC,OAAQ,EAAE,CAAC,OAAO,CAAC,2BAA2B,KAAK,GAAG,CAAC,CAAC,GAAG,EAAoB,CAAC,CAAC,CAAC;IAAC,CAAC;IAAC,MAAM,CAAC;QAAC,OAAO,CAAC,CAAC;IAAC,CAAC;AAChH,CAAC,CAAC;AAEF,MAAM,UAAU,cAAc,CAAC,OAAe,EAAE,OAAgB;IAC9D,MAAM,UAAU,GAAsB,EAAE,CAAC;IACzC,MAAM,MAAM,GAAsC,EAAE,CAAC;IACrD,MAAM,KAAK,GAAsC,EAAE,CAAC;IACpD,IAAI,KAAK,GAAa,EAAE,CAAC;IACzB,IAAI,CAAC;QAAC,KAAK,GAAG,WAAW,CAAC,OAAO,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC;IAAC,CAAC;IAAC,MAAM,CAAC,CAAC,oBAAoB,CAAC,CAAC;IACnG,MAAM,QAAQ,GAAG,OAAO,CAAC,CAAC,CAAC,gBAAgB,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;IAE5D,KAAK,MAAM,IAAI,IAAI,KAAK,CAAC,IAAI,EAAE,EAAE,CAAC;QAChC,MAAM,IAAI,GAAG,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC;QACtC,MAAM,KAAK,GAAG,gBAAgB,CAAC,IAAI,CAAC,CAAC;QACrC,IAAI,cAAc,CAAC,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC;YAAC,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC,CAAC;YAAC,SAAS;QAAC,CAAC;QACzE,IAAI,EAAE,GAA6B,IAAI,CAAC;QACxC,IAAI,CAAC;YACH,EAAE,GAAG,IAAI,QAAQ,CAAC,IAAI,EAAE,EAAE,QAAQ,EAAE,IAAI,EAAE,aAAa,EAAE,IAAI,EAAE,CAAC,CAAC;YACjE,MAAM,UAAU,GAAG,CAAC,CAAC,EAAE,CAAC,OAAO,CAAC,yEAAyE,CAAC,CAAC,GAAG,EAAE,CAAC;YACjH,IAAI,CAAC,UAAU,EAAE,CAAC;gBAAC,KAAK,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC,CAAC;gBAAC,SAAS;YAAC,CAAC;YAC3D,MAAM,IAAI,GAAG,EAAE,CAAC,OAAO,CAAC,wEAAwE,CAAC,CAAC,GAAG,EAClC,CAAC;YACpE,IAAI,QAAQ,IAAI,IAAI,KAAK,GAAG,QAAQ,KAAK;gBAAE,SAAS;YACpD,IAAI,SAAS,GAAkB,IAAI,CAAC;YACpC,IAAI,CAAC;gBACH,SAAS,GAAI,EAAE,CAAC,OAAO,CAAC,qEAAqE,CAAC,CAAC,GAAG,EAA2B,CAAC,CAAC,CAAC;YAClI,CAAC;YAAC,MAAM,CAAC,CAAC,oBAAoB,CAAC,CAAC;YAChC,UAAU,CAAC,IAAI,CAAC;gBACd,IAAI;gBACJ,OAAO,EAAE,IAAI,EAAE,QAAQ,IAAI,IAAI;gBAC/B,KAAK;gBACL,eAAe,EAAE,SAAS,CAAC,EAAE,EAAE,kBAAkB,CAAC;gBAClD,KAAK,EAAE,SAAS,CAAC,EAAE,EAAE,OAAO,CAAC;gBAC7B,KAAK,EAAE,SAAS,CAAC,EAAE,EAAE,OAAO,CAAC;gBAC7B,aAAa,EAAE,SAAS,CAAC,EAAE,EAAE,gBAAgB,CAAC;gBAC9C,QAAQ,EAAE,SAAS,CAAC,EAAE,EAAE,UAAU,CAAC;gBACnC,UAAU,EAAE,IAAI,EAAE,cAAc,IAAI,IAAI;gBACxC,SAAS;aACV,CAAC,CAAC;QACL,CAAC;QAAC,MAAM,CAAC;YACP,KAAK,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC,CAAC;QAC9B,CAAC;gBAAS,CAAC;YACT,EAAE,EAAE,KAAK,EAAE,CAAC;QACd,CAAC;IACH,CAAC;IAED,MAAM,YAAY,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,SAAS,CAAC,CAAC,CAAC;IAC5D,MAAM,UAAU,GAAG,UAAU,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,CAAC,CAAC;UAC1D,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,CAAC,CAAC;UACvC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,CAAC,CAAC;UACtC,YAAY,CAAC;IACjB,OAAO,EAAE,OAAO,EAAE,UAAU,EAAE,MAAM,EAAE,KAAK,EAAE,YAAY,EAAE,UAAU,EAAE,CAAC;AAC1E,CAAC;AAED,MAAM,CAAC,MAAM,QAAQ,GAAG,CAAC,CAAS,EAAU,EAAE;IAC5C,IAAI,CAAC,IAAI,IAAI,IAAI,CAAC;QAAE,OAAO,GAAG,CAAC,CAAC,GAAG,IAAI,IAAI,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,CAAC;IAC9D,IAAI,CAAC,IAAI,IAAI,IAAI,CAAC;QAAE,OAAO,GAAG,CAAC,CAAC,GAAG,IAAI,IAAI,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,CAAC;IAC9D,IAAI,CAAC,IAAI,IAAI;QAAE,OAAO,GAAG,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,CAAC;IACpD,OAAO,GAAG,CAAC,IAAI,CAAC;AAClB,CAAC,CAAC;AAUF,MAAM,UAAU,aAAa,CAC3B,OAAe,EACf,OAAe,EACf,MAA4D,EAC5D,OAAgB,EAChB,WAA2C;IAE3C,MAAM,IAAI,GAA4C,EAAE,MAAM,EAAE,OAAO,EAAE,CAAC;IAC1E,IAAI,WAAW,CAAC,MAAM,EAAE,CAAC;QACvB,OAAO;YACL,GAAG,IAAI,EAAE,EAAE,EAAE,KAAK,EAAE,MAAM,EAAE,EAAE,WAAW,EAAE;YAC3C,OAAO,EAAE,eAAe,MAAM,UAAU,WAAW,CAAC,MAAM,wBAAwB,WAAW,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,8FAA8F;SACrO,CAAC;IACJ,CAAC;IACD,MAAM,MAAM,GAAG,SAAS,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC;IAC3C,IAAI,CAAC,UAAU,CAAC,MAAM,CAAC,EAAE,CAAC;QACxB,OAAO,EAAE,GAAG,IAAI,EAAE,EAAE,EAAE,KAAK,EAAE,OAAO,EAAE,mBAAmB,OAAO,OAAO,MAAM,iBAAiB,MAAM,0CAA0C,EAAE,MAAM,EAAE,EAAE,MAAM,EAAE,EAAE,CAAC;IACvK,CAAC;IACD,MAAM,WAAW,GAAG,gBAAgB,CAAC,MAAM,CAAC,CAAC;IAE7C,IAAI,MAAM,KAAK,iBAAiB,EAAE,CAAC;QACjC,MAAM,OAAO,GAAG,cAAc,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC;QAC/D,IAAI,CAAC,OAAO,EAAE,CAAC;YACb,OAAO;gBACL,GAAG,IAAI,EAAE,EAAE,EAAE,KAAK,EAAE,MAAM,EAAE,EAAE,MAAM,EAAE,KAAK,EAAE,WAAW,EAAE,IAAI,EAAE,OAAO,EAAE;gBACzE,OAAO,EAAE,kGAAkG,MAAM,mBAAmB,QAAQ,CAAC,WAAW,CAAC,IAAI;oBAC3J,GAAG,OAAO,EAAE,eAAe,IAAI,GAAG,2BAA2B,OAAO,EAAE,KAAK,IAAI,GAAG,WAAW,OAAO,EAAE,KAAK,IAAI,GAAG,WAAW,OAAO,EAAE,aAAa,IAAI,GAAG,eAAe,OAAO,EAAE,QAAQ,IAAI,GAAG,aAAa;oBAC9M,sFAAsF;aACzF,CAAC;QACJ,CAAC;QACD,KAAK,MAAM,MAAM,IAAI,CAAC,EAAE,EAAE,MAAM,EAAE,MAAM,CAAC;YAAE,MAAM,CAAC,MAAM,GAAG,MAAM,EAAE,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;QACpF,OAAO,EAAE,GAAG,IAAI,EAAE,EAAE,EAAE,IAAI,EAAE,MAAM,EAAE,EAAE,MAAM,EAAE,UAAU,EAAE,WAAW,EAAE,EAAE,CAAC;IAC5E,CAAC;IAED,IAAI,MAAM,KAAK,QAAQ,EAAE,CAAC;QACxB,MAAM,EAAE,GAAG,IAAI,QAAQ,CAAC,MAAM,CAAC,CAAC;QAChC,IAAI,CAAC;YACH,EAAE,CAAC,MAAM,CAAC,0BAA0B,CAAC,CAAC;YACtC,EAAE,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;QACpB,CAAC;gBAAS,CAAC;YAAC,EAAE,CAAC,KAAK,EAAE,CAAC;QAAC,CAAC;QACzB,MAAM,UAAU,GAAG,gBAAgB,CAAC,MAAM,CAAC,CAAC;QAC5C,OAAO,EAAE,GAAG,IAAI,EAAE,EAAE,EAAE,IAAI,EAAE,MAAM,EAAE,EAAE,WAAW,EAAE,UAAU,EAAE,cAAc,EAAE,IAAI,CAAC,GAAG,CAAC,WAAW,GAAG,UAAU,EAAE,CAAC,CAAC,EAAE,EAAE,CAAC;IAC3H,CAAC;IAED,wFAAwF;IACxF,6EAA6E;IAC7E,MAAM,EAAE,GAAG,IAAI,QAAQ,CAAC,MAAM,CAAC,CAAC;IAChC,IAAI,CAAC;QACH,MAAM,QAAQ,GAAI,EAAE,CAAC,OAAO,CAAC,gEAAgE,WAAW,EAAE,CAAC,CAAC,GAAG,EAA2B,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC;QAC9J,MAAM,UAAU,GAAI,EAAE,CAAC,OAAO,CAAC,0HAA0H,WAAW,EAAE,CAAC,CAAC,GAAG,EAA6B,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC;QAC9N,MAAM,EAAE,GAAG,CAAC,CAAS,EAAU,EAAE,CAAC,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,EAAE,GAAG,EAAE,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,IAAI,CAAC;QACzF,MAAM,OAAO,GAAI,EAAE,CAAC,OAAO,CAAC,0DAA0D,EAAE,CAAC,QAAQ,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,QAAQ,CAAmB,CAAC,CAAC,CAAC;QACnJ,MAAM,WAAW,GAAI,EAAE,CAAC,OAAO,CAAC,wDAAwD,EAAE,CAAC,QAAQ,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,QAAQ,CAAmB,CAAC,CAAC,CAAC;QACrJ,MAAM,YAAY,GAAI,EAAE,CAAC,OAAO,CAAC,gEAAgE,EAAE,CAAC,UAAU,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,UAAU,CAAmB,CAAC,CAAC,CAAC;QAClK,IAAI,CAAC,OAAO,EAAE,CAAC;YACb,EAAE,CAAC,KAAK,EAAE,CAAC;YACX,OAAO;gBACL,GAAG,IAAI,EAAE,EAAE,EAAE,KAAK,EAAE,MAAM,EAAE,EAAE,OAAO,EAAE,WAAW,EAAE,YAAY,EAAE,UAAU,EAAE,WAAW,EAAE;gBAC3F,OAAO,EAAE,sGAAsG,OAAO,gBAAgB,WAAW,iBAAiB,YAAY,kCAAkC,WAAW,0GAA0G;aACtU,CAAC;QACJ,CAAC;QACD,EAAE,CAAC,WAAW,CAAC,GAAG,EAAE;YAClB,EAAE,CAAC,OAAO,CAAC,6CAA6C,EAAE,CAAC,QAAQ,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,QAAQ,CAAC,CAAC;YACjG,EAAE,CAAC,OAAO,CAAC,+CAA+C,EAAE,CAAC,QAAQ,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,QAAQ,CAAC,CAAC;YACnG,EAAE,CAAC,OAAO,CAAC,qDAAqD,EAAE,CAAC,UAAU,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,UAAU,CAAC,CAAC;QAC/G,CAAC,CAAC,EAAE,CAAC;QACL,EAAE,CAAC,MAAM,CAAC,0BAA0B,CAAC,CAAC;QACtC,EAAE,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;IACpB,CAAC;YAAS,CAAC;QACT,IAAI,EAAE,CAAC,IAAI;YAAE,EAAE,CAAC,KAAK,EAAE,CAAC;IAC1B,CAAC;IACD,MAAM,UAAU,GAAG,gBAAgB,CAAC,MAAM,CAAC,CAAC;IAC5C,OAAO;QACL,GAAG,IAAI,EAAE,EAAE,EAAE,IAAI;QACjB,MAAM,EAAE,EAAE,WAAW,EAAE,UAAU,EAAE,cAAc,EAAE,IAAI,CAAC,GAAG,CAAC,WAAW,GAAG,UAAU,EAAE,CAAC,CAAC,EAAE,UAAU,EAAE,WAAW,EAAE;KACpH,CAAC;AACJ,CAAC"}

package/dist/core/draftBrief.d.ts

@@ -0,0 +1,27 @@
+export interface DraftBrief {
+    url: string;
+    topQuery: string | null;
+    gap: string;
+    maxPassageScore: number | null;
+    voice: {
+        heading: string;
+        text: string;
+    }[];
+    facts: {
+        heading: string;
+        text: string;
+        relevance: number;
+    }[];
+    brief: string;
+}
+/**
+ * Assemble a GROUNDED writing brief for a page so the host model (Claude) can draft the fix in the
+ * author's own voice — no bundled LLM. The page's existing chunks supply both the voice exemplars
+ * (match tone/pace/lexicon) and the facts (the reranker picks the most query-relevant ones), so the
+ * draft is phrased from the site's own material, never invented. urlKey omitted → the weakest
+ * (lowest max_passage_score) page.
+ */
+export declare function buildDraftBrief(dataDir: string, siteUrl: string, urlKey?: string): Promise<DraftBrief | {
+    error: string;
+}>;
+//# sourceMappingURL=draftBrief.d.ts.map

package/dist/core/draftBrief.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"draftBrief.d.ts","sourceRoot":"","sources":["../../src/core/draftBrief.ts"],"names":[],"mappings":"AAKA,MAAM,WAAW,UAAU;IACzB,GAAG,EAAE,MAAM,CAAC;IACZ,QAAQ,EAAE,MAAM,GAAG,IAAI,CAAC;IACxB,GAAG,EAAE,MAAM,CAAC;IACZ,eAAe,EAAE,MAAM,GAAG,IAAI,CAAC;IAC/B,KAAK,EAAE;QAAE,OAAO,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;IAC3C,KAAK,EAAE;QAAE,OAAO,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAC;QAAC,SAAS,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;IAC9D,KAAK,EAAE,MAAM,CAAC;CACf;AAED;;;;;;GAMG;AACH,wBAAsB,eAAe,CACnC,OAAO,EAAE,MAAM,EACf,OAAO,EAAE,MAAM,EACf,MAAM,CAAC,EAAE,MAAM,GACd,OAAO,CAAC,UAAU,GAAG;IAAE,KAAK,EAAE,MAAM,CAAA;CAAE,CAAC,CAyDzC"}

package/dist/core/draftBrief.js

@@ -0,0 +1,69 @@
+import { AuditDatabase } from './AuditDatabase.js';
+import { dbPathFor } from './paths.js';
+import { gscFreshness } from './gscFreshness.js';
+import { scorePassages } from './reranker.js';
+/**
+ * Assemble a GROUNDED writing brief for a page so the host model (Claude) can draft the fix in the
+ * author's own voice — no bundled LLM. The page's existing chunks supply both the voice exemplars
+ * (match tone/pace/lexicon) and the facts (the reranker picks the most query-relevant ones), so the
+ * draft is phrased from the site's own material, never invented. urlKey omitted → the weakest
+ * (lowest max_passage_score) page.
+ */
+export async function buildDraftBrief(dataDir, siteUrl, urlKey) {
+    const db = new AuditDatabase(dbPathFor(dataDir, siteUrl));
+    try {
+        const sel = 'SELECT url_key url, body_chunks bc, max_passage_score mps, max_passage_query mpq, title FROM pages';
+        const page = (urlKey
+            ? db.db.prepare(`${sel} WHERE url_key = ?`).get(urlKey)
+            : db.db.prepare(`${sel} WHERE max_passage_score IS NOT NULL AND body_chunks IS NOT NULL ORDER BY max_passage_score ASC LIMIT 1`).get());
+        if (!page)
+            return { error: `No such page${urlKey ? ` ${urlKey}` : ''}.` };
+        if (!page.bc)
+            return { error: 'Page has no body_chunks — run start_crawl (then score_passages) first.' };
+        let chunks;
+        try {
+            chunks = JSON.parse(page.bc);
+        }
+        catch {
+            return { error: 'Could not parse body_chunks.' };
+        }
+        chunks = chunks.filter(c => c.text && c.text.length > 40);
+        // Top query: the one the page was scored against; else the highest-impression GSC query.
+        let topQuery = page.mpq;
+        if (!topQuery) {
+            const maxDate = gscFreshness(db.db).effectiveMax;
+            if (maxDate) {
+                const q = db.db.prepare(`SELECT query FROM search_analytics WHERE page_key = ? AND query IS NOT NULL AND date <= '${maxDate}' AND date > date('${maxDate}','-28 days') GROUP BY query ORDER BY SUM(impressions) DESC LIMIT 1`).get(page.url);
+                topQuery = q?.query ?? null;
+            }
+        }
+        // Grounding facts — rerank the page's own chunks against the query, keep the most relevant.
+        let facts = [];
+        if (topQuery && chunks.length) {
+            const passages = chunks.map(c => `${c.heading ? `${c.heading}. ` : ''}${c.text}`);
+            const scores = await scorePassages(topQuery, passages);
+            facts = chunks.map((c, i) => ({ heading: c.heading ?? '', text: (c.text ?? '').slice(0, 500), relevance: Math.round((scores[i] ?? 0) * 100) / 100 }))
+                .sort((a, b) => b.relevance - a.relevance).slice(0, 4);
+        }
+        else {
+            facts = chunks.slice(0, 4).map(c => ({ heading: c.heading ?? '', text: (c.text ?? '').slice(0, 500), relevance: 0 }));
+        }
+        // Voice exemplars — a couple of the page's substantial passages (tone/pace/lexicon to match).
+        const voice = chunks.filter(c => (c.text ?? '').length > 120).slice(0, 2).map(c => ({ heading: c.heading ?? '', text: (c.text ?? '').slice(0, 450) }));
+        const gap = page.mps == null ? 'not scored (run score_passages)'
+            : page.mps < 3 ? `weak-passage (best passage scores ${page.mps} — no section confidently answers the query)`
+                : `the page already answers this (score ${page.mps}) — consider a refresh/expansion instead`;
+        const brief = [
+            `Make the SMALLEST change that closes the gap for the query “${topQuery ?? '(unknown)'}”: draft ONE self-contained passage to slot in (a clear heading + ~50–90 words) — not a rewrite. Preserve the rest of the page; this is a surgical addition (less intervention is better).`,
+            `Lead with a one-sentence direct answer, then support it.`,
+            `VOICE: match the tone, pace and lexicon of the voice samples exactly — first-person, from someone who owns and uses the gear, naming specific products. British English, contractions, no tidy "in conclusion" bow.`,
+            `AVOID AI slop: no em dashes; no clichés ("landscape", "elevate", "whether you're…", "in today's…", "dive in", "it's worth noting"); no formal-transition padding.`,
+            `GROUND STRICTLY: use only facts present in the facts/voice content below — invent no specs, prices, dates or claims. If a needed fact is missing, write around what's there and flag the gap rather than guessing.`,
+        ].join(' ');
+        return { url: page.url, topQuery, gap, maxPassageScore: page.mps, voice, facts, brief };
+    }
+    finally {
+        db.close();
+    }
+}
+//# sourceMappingURL=draftBrief.js.map

package/dist/core/draftBrief.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"draftBrief.js","sourceRoot":"","sources":["../../src/core/draftBrief.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,aAAa,EAAE,MAAM,oBAAoB,CAAC;AACnD,OAAO,EAAE,SAAS,EAAE,MAAM,YAAY,CAAC;AACvC,OAAO,EAAE,YAAY,EAAE,MAAM,mBAAmB,CAAC;AACjD,OAAO,EAAE,aAAa,EAAE,MAAM,eAAe,CAAC;AAY9C;;;;;;GAMG;AACH,MAAM,CAAC,KAAK,UAAU,eAAe,CACnC,OAAe,EACf,OAAe,EACf,MAAe;IAEf,MAAM,EAAE,GAAG,IAAI,aAAa,CAAC,SAAS,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC,CAAC;IAC1D,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,oGAAoG,CAAC;QACjH,MAAM,IAAI,GAAG,CAAC,MAAM;YAClB,CAAC,CAAC,EAAE,CAAC,EAAE,CAAC,OAAO,CAAC,GAAG,GAAG,oBAAoB,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC;YACvD,CAAC,CAAC,EAAE,CAAC,EAAE,CAAC,OAAO,CAAC,GAAG,GAAG,yGAAyG,CAAC,CAAC,GAAG,EAAE,CACvB,CAAC;QAClH,IAAI,CAAC,IAAI;YAAE,OAAO,EAAE,KAAK,EAAE,eAAe,MAAM,CAAC,CAAC,CAAC,IAAI,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,GAAG,EAAE,CAAC;QAC1E,IAAI,CAAC,IAAI,CAAC,EAAE;YAAE,OAAO,EAAE,KAAK,EAAE,wEAAwE,EAAE,CAAC;QAEzG,IAAI,MAA6C,CAAC;QAClD,IAAI,CAAC;YAAC,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QAAC,CAAC;QAAC,MAAM,CAAC;YAAC,OAAO,EAAE,KAAK,EAAE,8BAA8B,EAAE,CAAC;QAAC,CAAC;QACjG,MAAM,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,IAAI,CAAC,MAAM,GAAG,EAAE,CAAC,CAAC;QAE1D,yFAAyF;QACzF,IAAI,QAAQ,GAAG,IAAI,CAAC,GAAG,CAAC;QACxB,IAAI,CAAC,QAAQ,EAAE,CAAC;YACd,MAAM,OAAO,GAAG,YAAY,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC,YAAY,CAAC;YACjD,IAAI,OAAO,EAAE,CAAC;gBACZ,MAAM,CAAC,GAAG,EAAE,CAAC,EAAE,CAAC,OAAO,CACrB,4FAA4F,OAAO,sBAAsB,OAAO,qEAAqE,CACtM,CAAC,GAAG,CAAC,IAAI,CAAC,GAAG,CAAkC,CAAC;gBACjD,QAAQ,GAAG,CAAC,EAAE,KAAK,IAAI,IAAI,CAAC;YAC9B,CAAC;QACH,CAAC;QAED,4FAA4F;QAC5F,IAAI,KAAK,GAAwB,EAAE,CAAC;QACpC,IAAI,QAAQ,IAAI,MAAM,CAAC,MAAM,EAAE,CAAC;YAC9B,MAAM,QAAQ,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,OAAO,IAAI,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;YAClF,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC,QAAQ,EAAE,QAAQ,CAAC,CAAC;YACvD,KAAK,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,EAAE,OAAO,EAAE,CAAC,CAAC,OAAO,IAAI,EAAE,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,SAAS,EAAE,IAAI,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,GAAG,EAAE,CAAC,CAAC;iBAClJ,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,SAAS,GAAG,CAAC,CAAC,SAAS,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;QAC3D,CAAC;aAAM,CAAC;YACN,KAAK,GAAG,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,EAAE,OAAO,EAAE,CAAC,CAAC,OAAO,IAAI,EAAE,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,SAAS,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC;QACxH,CAAC;QAED,8FAA8F;QAC9F,MAAM,KAAK,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC,MAAM,GAAG,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,EAAE,OAAO,EAAE,CAAC,CAAC,OAAO,IAAI,EAAE,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC;QAEvJ,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,IAAI,IAAI,CAAC,CAAC,CAAC,iCAAiC;YAC9D,CAAC,CAAC,IAAI,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC,CAAC,qCAAqC,IAAI,CAAC,GAAG,8CAA8C;gBAC1G,CAAC,CAAC,wCAAwC,IAAI,CAAC,GAAG,0CAA0C,CAAC;QAEjG,MAAM,KAAK,GAAG;YACZ,+DAA+D,QAAQ,IAAI,WAAW,4LAA4L;YAClR,0DAA0D;YAC1D,qNAAqN;YACrN,mKAAmK;YACnK,oNAAoN;SACrN,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAEZ,OAAO,EAAE,GAAG,EAAE,IAAI,CAAC,GAAG,EAAE,QAAQ,EAAE,GAAG,EAAE,eAAe,EAAE,IAAI,CAAC,GAAG,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,CAAC;IAC1F,CAAC;YAAS,CAAC;QACT,EAAE,CAAC,KAAK,EAAE,CAAC;IACb,CAAC;AACH,CAAC"}

package/dist/core/extract.d.ts

@@ -0,0 +1,67 @@
+import { type UrlKeyOptions } from './url-key.js';
+/**
+ * Internal = the seed host or its www/apex twin ONLY — not arbitrary sibling subdomains.
+ * accounts./shop./blog. etc. are separate properties; crawling (or following redirects into)
+ * them pulls in junk like Shopify's accounts.<domain> OAuth login pages. Shared by the
+ * link-classifier here and the crawler's redirect-target guard.
+ */
+export declare const isInternalHost: (targetHost: string, baseHost: string) => boolean;
+export interface ExtractedLink {
+    targetUrl: string;
+    targetKey: string;
+    anchor: string;
+    placement: 'navigation' | 'footer' | 'body';
+    rel: string;
+    isInternal: boolean;
+}
+export interface ExtractedPage {
+    title: string | null;
+    titleLength: number;
+    metaDescription: string | null;
+    metaDescriptionLength: number;
+    h1: string | null;
+    h1Count: number;
+    wordCount: number;
+    lang: string | null;
+    charset: string | null;
+    canonicalUrl: string | null;
+    canonicalKey: string | null;
+    robots: string | null;
+    viewport: string | null;
+    noindex: boolean;
+    jsonLd: string | null;
+    ogTags: string | null;
+    hreflang: string | null;
+    internalLinks: number;
+    externalLinks: number;
+    imageCount: number;
+    imagesWithoutAlt: number;
+    imagesMissingDimensions: number;
+    canonicalCount: number;
+    canonicalRelative: boolean;
+    h2Count: number;
+    headingSkips: number;
+    relNext: boolean;
+    relPrev: boolean;
+    relAmphtml: string | null;
+    mixedContentCount: number;
+    twitterTags: string | null;
+    hasMicrodata: boolean;
+    hasRdfa: boolean;
+    hasFavicon: boolean;
+    hasAnalytics: boolean;
+    bodyChunks: {
+        heading: string;
+        level: number;
+        text: string;
+    }[];
+    links: ExtractedLink[];
+    imageSrcs: string[];
+}
+/**
+ * Extract the SEO-relevant surface from raw HTML. Pure function of bytes (D).
+ * `baseHost` is the seed apex host for internal/external classification;
+ * `keyOpts` is the url-key normalisation config for this property.
+ */
+export declare function extractPage(html: string, pageUrl: string, baseHost: string, keyOpts: UrlKeyOptions, xRobotsTag?: string | null): ExtractedPage;
+//# sourceMappingURL=extract.d.ts.map

package/dist/core/extract.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"extract.d.ts","sourceRoot":"","sources":["../../src/core/extract.ts"],"names":[],"mappings":"AACA,OAAO,EAAU,KAAK,aAAa,EAAE,MAAM,cAAc,CAAC;AAE1D;;;;;GAKG;AACH,eAAO,MAAM,cAAc,GAAI,YAAY,MAAM,EAAE,UAAU,MAAM,KAAG,OAKrE,CAAC;AAEF,MAAM,WAAW,aAAa;IAC5B,SAAS,EAAE,MAAM,CAAC;IAClB,SAAS,EAAE,MAAM,CAAC;IAClB,MAAM,EAAE,MAAM,CAAC;IACf,SAAS,EAAE,YAAY,GAAG,QAAQ,GAAG,MAAM,CAAC;IAC5C,GAAG,EAAE,MAAM,CAAC;IACZ,UAAU,EAAE,OAAO,CAAC;CACrB;AAED,MAAM,WAAW,aAAa;IAC5B,KAAK,EAAE,MAAM,GAAG,IAAI,CAAC;IACrB,WAAW,EAAE,MAAM,CAAC;IACpB,eAAe,EAAE,MAAM,GAAG,IAAI,CAAC;IAC/B,qBAAqB,EAAE,MAAM,CAAC;IAC9B,EAAE,EAAE,MAAM,GAAG,IAAI,CAAC;IAClB,OAAO,EAAE,MAAM,CAAC;IAChB,SAAS,EAAE,MAAM,CAAC;IAClB,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;IACpB,OAAO,EAAE,MAAM,GAAG,IAAI,CAAC;IACvB,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,MAAM,EAAE,MAAM,GAAG,IAAI,CAAC;IACtB,QAAQ,EAAE,MAAM,GAAG,IAAI,CAAC;IACxB,OAAO,EAAE,OAAO,CAAC;IACjB,MAAM,EAAE,MAAM,GAAG,IAAI,CAAC;IACtB,MAAM,EAAE,MAAM,GAAG,IAAI,CAAC;IACtB,QAAQ,EAAE,MAAM,GAAG,IAAI,CAAC;IACxB,aAAa,EAAE,MAAM,CAAC;IACtB,aAAa,EAAE,MAAM,CAAC;IACtB,UAAU,EAAE,MAAM,CAAC;IACnB,gBAAgB,EAAE,MAAM,CAAC;IACzB,uBAAuB,EAAE,MAAM,CAAC;IAChC,cAAc,EAAE,MAAM,CAAC;IACvB,iBAAiB,EAAE,OAAO,CAAC;IAC3B,OAAO,EAAE,MAAM,CAAC;IAChB,YAAY,EAAE,MAAM,CAAC;IACrB,OAAO,EAAE,OAAO,CAAC;IACjB,OAAO,EAAE,OAAO,CAAC;IACjB,UAAU,EAAE,MAAM,GAAG,IAAI,CAAC;IAC1B,iBAAiB,EAAE,MAAM,CAAC;IAC1B,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,YAAY,EAAE,OAAO,CAAC;IACtB,OAAO,EAAE,OAAO,CAAC;IACjB,UAAU,EAAE,OAAO,CAAC;IACpB,YAAY,EAAE,OAAO,CAAC;IACtB,UAAU,EAAE;QAAE,OAAO,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;IAC/D,KAAK,EAAE,aAAa,EAAE,CAAC;IACvB,SAAS,EAAE,MAAM,EAAE,CAAC;CACrB;AASD;;;;GAIG;AACH,wBAAgB,WAAW,CACzB,IAAI,EAAE,MAAM,EACZ,OAAO,EAAE,MAAM,EACf,QAAQ,EAAE,MAAM,EAChB,OAAO,EAAE,aAAa,EACtB,UAAU,CAAC,EAAE,MAAM,GAAG,IAAI,GACzB,aAAa,CA2Nf"}